Head and neck cancers have well-documented associations with tobacco and alcohol exposure, but the disease develops in only a small fraction of users, which implies an important role for genetic susceptibility. Therefore, head and neck cancers are an excellent model for studying genetic susceptibility to environmental carcinogens. The primary goal of this R01 application is to perform a comprehensive two-stage, high-density, genome-wide single-nucleotide polymorphism (SNP) analysis of head and neck cancer cases and corresponding frequency matched controls to identify novel genetic risk factors for head and neck cancer. This proposal builds upon a well-annotated existing DNA repository of cases and controls. One of the unique features of our study is the availability of DNA repair assay data on most of the cases and controls in this study, which will allow us to conduct genotype/phenotype analyses. We also have access to genome-wide association data from 1200 white control subjects from the same source population. In aim 1, we will perform genotyping on 1000 randomly selected head and neck cancer cases and 500 controls using a 370K Illumina Infinium HapMap HumanCNV370-Duo SNP Chip. We will perform association analyses (1000 cases and 1700 controls) in the first stage using outcome variable as case-control status as well as DNA repair capacity assay data. Our second aim is to perform second-stage analysis of the SNPs selected in stage 1 using 900 additional cases and corresponding controls from the same source and from UCLA. We will use efficient joint analysis of cases and controls from the first and second aims, for a total of 1900 cases and 2600 controls. Finally, in aim 3, we will apply novel statistical tools such as the latent variable approach with Tukey's one-degree-of-freedom test and support vector machines to identify gene-gene and gene-environment (using environmental factors such as smoking and alcohol use) interactions that contribute to the risk of head and neck cancer. We are an experienced investigative team proposing a comprehensive analysis that incorporates epidemiological, behavioral, and functional data. Identification of novel genetic risk factors and their interactions with environmental factors will contribute to the early diagnosis of head and neck cancers.